10.9/The Interior Life/The Private Life

Ordinary Violence: Petty Crime, Harassment, and the Managed Risk, 1989–2001

The Stop With a Legal Name

That was one accounting of the decade’s ordinary danger. The other began with a number. New York City recorded 2,245 murders in 1990, the highest tally in the city’s recorded history — a rate of a little over six killings a day, citywide.³ A New Yorker walking to the office that same year carried, as often as not, two wallets. The real one rode in a front trouser pocket or deep inside a handbag held close against the body. The other sat wherever a hand reaching from behind found it fastest: twenty or thirty dollars, an expired credit card, sometimes an old library card — the props of a wallet rather than the thing itself. New Yorkers called it mugger money. Handed over quickly enough, it ended an encounter before it became something worse; a folded five-dollar bill kept in a shoe or a sock was the fallback if the mugger took the rest. None of it registered as unusual. It was routine preparation for a city in which being robbed at knife- or gunpoint was understood as a matter of when, not whether.

The habit came with a larger, unwritten set of rules. Automated teller machines were used before nine at night and never inside an enclosed vestibule, which through the first half of the 1990s — before New York’s 1996 ATM Safety Act required mirrors and cameras at the machine — carried no such equipment and was understood, on its own terms, as the more dangerous of the two locations. Cash withdrawn at a machine went straight into a pocket, counted later, indoors, never on the sidewalk. Certain sneakers were a known robbery target for teenagers in some neighborhoods, and a Walkman worn with its cord visible read less as convenience than as invitation. After dark, a pedestrian walked in the street rather than past a row of recessed doorways, and toward oncoming traffic rather than with its flow; someone who thought they were being followed ducked into the nearest open business rather than continuing home. The same era logged more than 150,000 stolen cars a year citywide, roughly one every three minutes, though the arrest figures on record — 16,600 in 1989 — counted only the thieves who got caught, a small fraction of the total.⁴ The mental map this arithmetic produced was built from more than the murder count alone — from visible drug activity, from graffiti, from the ordinary unease of a dark block — and it did not always track the official numbers with much precision, so that a resident could feel considerably more exposed than the year’s own homicide tally, taken block by block, would have justified. None of it was posted anywhere. It passed from person to person the way local weather did, refreshed by whoever had just been mugged and was willing to say so out loud.

The Rules No One Wrote Down

“Sexual harassment” existed by 1989 as a legal concept, but only inside the workplace, following the Supreme Court’s 1986 ruling in Meritor Savings Bank v. Vinson that a hostile environment could itself violate federal civil-rights law.⁵ Nothing carried that concept onto the subway. A hand on a crowded train, a man pressed against a rider’s back on a rush-hour local, a man exposing himself against a platform pillar or in an empty car — none of it had a name used in ordinary conversation. Transit felonies alone ran to 17,497 in 1990, close to forty-eight a day system-wide, and the transit police force assigned to a system that size had neither the staffing nor, by most accounts, the inclination to treat the subway’s routine harassment as a category of crime worth its own resources.⁶ The Metropolitan Transportation Authority kept no formal channel for reporting it. A woman who mentioned what had happened to her was as likely to be told to ignore it as to be believed, and most women, reasonably, stopped mentioning it.

What existed instead was a set of rules, passed from woman to woman and never written down anywhere an agency would have filed it. Which car to board, where to stand on the platform, when the arithmetic of a cab fare beat the arithmetic of what might happen on the last train — all of it functioned as a private civic infrastructure running underneath the one the Transit Authority published on its maps. Lines built to the old IRT standard had, among riders, a worse reputation than the IND lines built later, and a train that ran through a stretch of the Bronx or Upper Manhattan a rider knew well read as safer, in practice, than an unfamiliar line through an unfamiliar neighborhood, however the two compared on paper.

William Bratton’s arrival as chief of the Transit Police in 1990 brought a fare-evasion crackdown that cut the transit felony count by roughly fifteen percent within its first year, and the subway continued to feel measurably different by the middle of the decade — not safe, exactly, but no longer immediately dangerous.⁷ The improvement was not evenly shared: a rider changing trains at a well-lit Midtown station after a fare crackdown felt the change more than a rider standing on a quiet outer-borough platform at midnight, where the additional police presence rarely reached. And the crackdown, in either case, did nothing to give the older, unwritten problem — the hand on the crowded car, the man against the pillar — a public vocabulary, a report form, or an office to complain to.

The Gauntlet

Midtown Manhattan ran under scaffolding for most of the period — sidewalk sheds thrown up for construction and never entirely taken down before the next building needed one. A woman leaving an office at the end of the day walked the length of one on the way to the subway or the bus, past workers positioned above and at street level, and the commentary that came down from the scaffold ran a known sequence: a greeting, then an escalation if the greeting went unanswered, the silence itself recast as an offense requiring a sharper remark. Being followed on the sidewalk in broad daylight drew no attention from anyone else on the block; a car slowing to keep pace with a pedestrian was common enough in parts of the outer boroughs to warrant no comment at all. None of it had an organizing name. What existed as vocabulary belonged to men — “catcalling” was already casual slang for the whistle and the comment — but no word treated the daily accumulation of it as a subject a woman could formally raise, and the options when it happened were the same three that had always existed: answer, which risked escalation; ignore, which felt like surrender; or mention it afterward to a colleague who had walked the same block and already knew.

Geography sorted who absorbed how much of it, and the sorting followed class as much as neighborhood. A woman moving between a doorman building and a car service met the gauntlet rarely, if at all; a woman walking to a subway stop from a walk-up apartment met it daily, on foot, both ways, through a Midtown corridor thick with scaffolding on the way in and the way home. In a gentrifying downtown block where an old population and a new one occupied the same sidewalk without yet agreeing on its rules, a woman’s presence read ambiguously to men who had lived there longer, and the calculus of who was safe to ignore a comment from ran differently than it did in a neighborhood built entirely of newcomers. Farther out, in blocks where the man on the stoop was a known face rather than a stranger, the harassment took a more personal, less anonymous form — recognizable, sometimes, as a specific neighbor rather than as weather — which changed what a woman could say about it afterward, and to whom. The magazine’s own staff sat across all of these divides at once without ever discussing them as one problem.

The Wall Between Apartments

Prewar apartment buildings did not seal sound the way later construction would. Plaster walls carried voices; radiator risers carried them further. A building’s residents knew a shouting match from ordinary argument, and they knew the different acoustic signature of a blow landing, and they knew, in buildings where it recurred, roughly which night of the week to expect it and which door it came from. Almost none of them called the police.

The calculation against calling was not paranoia. An officer arriving at a domestic dispute in 1989 worked from a professional culture, formalized in an influential 1967 training manual from the International Association of Chiefs of Police, that treated arrest as a last resort — the standard response was to separate the parties, tell the aggressor to take a walk, and leave absent visible injury.⁸ A victim who went to Family Court instead met a jurisdictional wall of its own: under Article 8 of the Family Court Act as it stood through 1994, pursuing a Family Court order of protection and a criminal prosecution at the same time was, in practice, difficult to do at once, and a Family Court order carried no automatic alert to the officers who might later answer a call at that address. In the Bronx alone, close to eight thousand women filed Family Court protection petitions in 1989, a volume the system’s dual-track structure was not built to resolve quickly.⁹ Calling 911 could also make things worse rather than better — an abuser who learned someone had called was an abuser with a fresh grievance — and in neighborhoods where the police themselves carried a separate reputation for harm, a call for help introduced a second variable rather than removing the first.

The federal Violence Against Women Act, signed into law on September 13, 1994, was the first federal statute to treat domestic violence and sexual assault explicitly as crimes warranting dedicated prosecution funding.¹⁰ The Police Department’s own thinking had already begun shifting that year: a departmental strategy paper titled “Breaking the Cycle of Domestic Violence,” issued in 1994, argued internally for treating a family dispute as a police matter rather than a private one before state law required it to.¹¹ New York’s own Family Protection and Domestic Violence Intervention Act, passed the same year, eliminated the Family Court Act’s choice-of-forum barrier outright; its mandatory-arrest provision, requiring an officer to arrest on probable cause in a family-offense case regardless of the victim’s stated preference, took effect on July 1, 1995.¹² For the five years before that date, the wall between apartments carried what it carried, and the people on either side of it made their own arrangements with what the law did not yet do for them.

The Precinct That Solved Its Own Crimes

Mayor David Dinkins established the Mollen Commission by executive order in July 1992, naming the former deputy mayor and appellate judge Milton Mollen to chair it, after the May 1992 arrest of a veteran narcotics detective assigned to the 75th Precinct in East New York, Brooklyn, on charges of running a drug and robbery operation out of his own patrol car. The Police Department’s Internal Affairs Bureau had received complaints about the detective as early as 1986 and had not acted on them. The commission’s report, released July 7, 1994, found corruption concentrated in roughly ten precincts — among them the 75th, the 30th in Harlem, the 48th in the Bronx, the 73rd in Brownsville — all high-crime, all with an active narcotics trade, all in predominantly Black or Latino neighborhoods, and found “willful blindness” running through the department’s own ranks: dozens of prior corruption cases involving senior officers, buried rather than pursued.¹³ A sergeant who had supervised roughly seven hundred fifty homicide investigations out of the 75th Precinct during those years testified to the commission that only one, to his knowledge, had been conducted correctly.¹³

The 30th Precinct supplied the largest single case. On September 28, 1994, twenty-nine of its officers were arrested in a single day, on charges running from perjury and extortion to grand larceny and narcotics distribution; thirty-three were eventually charged, in what stood at the time as the largest corruption case brought against the department in nearly a decade.¹⁴ The specific method the precinct’s own officers called “booming” — filing a false radio call to justify an illegal search of a suspected dealer’s apartment, then keeping the drugs and cash the search turned up rather than logging them into evidence — meant that officer testimony had put people in prison on searches that had never happened as described. A hundred and twenty-five cases against ninety-eight defendants were dismissed once the pattern came to light, the largest single perjury-driven dismissal on record in the state to that point.¹⁵

For residents of the neighborhoods those precincts served, the finding confirmed something many had already concluded from direct experience: that a call to 911 might bring help, or might bring an officer who had planted evidence on a neighbor’s son the year before, and that the two possibilities were not reliably distinguishable in advance. For readers of the newspaper coverage the report generated on the Upper East Side or the Upper West Side, it read differently — as corruption confined to precincts other than their own, confirming a geography of policing they had already assumed rather than disturbing it.

The Count and Its Price

Between 1990 and the century’s end, murders in New York City fell from 2,245 to the low six hundreds, robberies from just over one hundred thousand to under twenty-eight thousand — a decline City Hall credited to the CompStat management system introduced in 1994 and the aggressive low-level enforcement, popularly called broken-windows policing, that accompanied it.⁶ The credit was contested from the start. Researchers pointed out that the decline had begun before 1994, under the prior administration, and that the crack cocaine market whose spread had driven much of the murder count of the late 1980s was, on its own internal timeline, contracting — its regular users aging out, its open-air violence over territory consolidating as fewer new buyers entered the market.¹⁶ That contraction was already pulling the homicide count downward before CompStat’s first meeting, a sequence documented in later book-length studies of the decline.¹⁷ Of 4,298 homicide victims studied citywide across 1990 and 1991, eighty-seven percent were Black or Latino, and investigators attributed most of that violence to the drug trade already treated on its own terms elsewhere in this archive.¹⁸

Whichever explanation carried the greater share, the enforcement side of the decade’s answer had its own accounting. Quality-of-life enforcement — arrests for fare-beating, public drinking, public urination, the kind of offense a broken-windows strategy treated as a precursor to worse crime — left behind a paper trail of criminal records that followed the people who accumulated them into job applications, housing applications, and custody hearings for years afterward, and that trail, like the enforcement generating it, fell across the city unevenly, concentrated in the same neighborhoods where the stop-and-frisk described at the start of this chapter had already made itself a weekly fact of life. The drug markets that the decade’s enforcement targeted did not disappear so much as move and grow quieter, a subject with its own accounting elsewhere in this archive. The cost of the count found its endpoint in the killing of Amadou Diallo, an unarmed West African immigrant shot in his own Bronx doorway in February 1999 — an event whose fuller account belongs to this archive’s record of the city’s crime geography. The safety a falling murder count bought was not distributed evenly across the population that paid for it.